The purpose of this proposal is to develop a performance theory to account for aggression which occurs in family interactions. Specifically we will explore various applications of the negative reinforcement paradigms as it might apply to a performance theory. The literature reviewed suggests several omissions crucial to this application. A combination of laboratory experiments, manipulations in field settings, and analyses of sequential interactions in the home are outlined. Three different procedures are designed to contribute to the areas which require further investigation. These general areas are: 1. in the natural setting, does termination of aversive inputs lead to a termination of aggression on the part of the person attacked? 2. Given consistent negative reinforcement arrangements, is there an across-trials increase in the probability of coercive responses being used to remove aversive inputs? 3. Does a negative reinforcement arrangement contingent upon intensity result in across-trials escalation in intensity? 4. Can it be demonstrated experimentally that power imbalance will disrupt such escalation, i.e., given that one person employs a higher amplitude aversive stimulus? 5. Does power imbalance between mother and problem child covary with extended aversive interchanges? 6. How much awareness does a parent have who is involved in negative reinforcement arrangements? Throughout the proposed research, specific emphasis will be placed upon a understanding of the coercion process as it relates to family interaction. It is assumed that an increased understanding would have immediate implications for both treatment and prevention.